by Ben Brothers, 1999
Two of the great technologies of the twentieth century are the computer and long-distance communication. Their convergence into the Internet represents one of the fundamental revolutions of our time. But the Internet began as a modest project, with a goal that is completely unrelated to that of global connectivity.
This fact, that the Internet has evolved and changed drastically since its conception, is the source of a great deal of interest, particularly in those camps which would argue that technology is out of control, or that mankind has been overrun by his machines. This is not a philosophy that I find very attractive, and more importantly, it is not a philosophy that I find very true.
The same argument is often applied to the transistor, with similar results. The revolution the transistor has sparked in the electronic world is undeniable. Its inventors, John Bardeen, Walter Brattain and William Schockley received the Nobel Prize in Physics in 1956. But even at that point, its impact was not fully realized.
When the transistor was invented, in 1947, it was designed as an amplifier for cross-country telephone signals. It was only later that the transistor's secondary property, that of switching, was capitalized on. The CMOS (complementary metal-oxide semiconductor) transistor (which has no steady-state current flow, and thus, no steady-state current), has made possible, via logic gates composed of switches, the micro-controller and the modern computer.
The invention and evolution of the Internet is similar in many ways to the invention and evolution of the transistor, and both do a good job of highlighting how technologies grow and mature. Both have evolved far beyond their original design, and have proved far more beneficial to mankind than would have seemed possible fifty years ago.
One popular train of thought suggests that these inventions have, more or less a life of their own. Proponents of this theory often make claims such as "technologies may develop independently of human agency". Another, more popular idea suggests that technologies evolve, but that "human choice influences the course of technological development."
I call the first idea "strong technological determinism", and the second idea "weak technological determinism". Both of these theories seem to me to suggest that things developed the way the did, because conditions at the time forced them to. If Mr. Bardeen and Mr. Brattain hadn't invented the transistor, someone else would have, at about the same time. If Leonard Kleinrock hadn't invented the theory of packet transmission, someone else would have, at about the same time. If Einstein hadn't developed the theory of general relativity, someone else would have, at about the same time.
This is, at its worst, merely a form of historical determinism. At its best, it is an insult to the hard work of thousands of scientists and engineers who thought about things in new ways, and turned new ideas into new inventions. Plutarch once observed that the mind is not a vessel to be filled, but a fire to be lighted. To deny that these fires exist, or at least to deny that they really matter, in any substantial way, seems to me the logical conclusion of such a philosophy.
I suggest that human choice does not merely "influence" technological development. I suggest that human choice is the sole engine of technological development. Every design, every experiment, every idea, is the direct result of individual thought. Of course, ideas feed on each other, and we can pool our resources in pursuit of the best one. This is one meaning of Newton's remark that "if I have seen further than others, it is because I have stood on the shoulders of giants". (Newton was also subtly insulting a vertically-challenged colleague with whom he often disagreed.) But we cannot forget or ignore the role of the individual.
Our discussion here concerns the development of the Internet, and here also we must avoid the temptation to ascribe innovation to the circumstantial. We need to remember that every RFC, every protocol, every transmission program, was designed by someone who had a unique insight into a problem. Technologies are not alive. They do not evolve on their own, like biological creatures. They evolve when men find new uses for them, and adapt them to those uses.
In 1961, Leonard Kleinrock, a professor at MIT, published a paper entitled "Information Flow in Large Communication Nets". It was the first paper to introduce the theory of packet-switching, which would later become the basis for the governing Internet protocols.
Later in the decade, the RAND Corporation, charged with the task of protecting America's communication grid from nuclear attack, conceived a distributed network system with no central control. The system was assumed to be entirely unreliable; every data packet had to successfully navigate the network by itself. Additionaly, every node on the network would be able to send, receive and tranmit data packets. Intitial testing of this network was done by RAND, MIT and UCLA, and the first test network was set up in 1968 at the National Physical Laboratory in Great Britain. In the US, the Advanced Research Projects Agency (ARPA), began construction of a more ambitious network in 1969; by December of that year, there were four nodes on the new ARPANET.
By 1971, 15 nodes were online and by 1972, there were 37. It was at the same time that RFC 318 outlined the specifications for telnet, and Ray Tomlinson of BBN wrote the first true email program (which included such familiar features as the @ sign).
By this point it was clear that research and science had usurped the ARPANET's original intent of defensive communication. Researchers using America's supercomputers (which were still at this time very rare) gained valuable time and insight by networking their programs and sharing their results.
As the number of users grew, it became clear that the Network Control Protocol (NCP) which governed the ARPANET was in need of replacing. The solution, TCP/IP was true to the ARPANET's decentralized nature. TCP/IP is actually a dual system, consisting of TCP, the transmission control protocol, and IP, the Internet protocol. TCP os responsible for configuring the data packets at the source, and reconfiguring them at the destination. TCP is totally blind with respect to the network itself. Similarly, IP is soley responsible for routing the packets from source to destination, across any arbitrary path. IP knows nothing about the actual data it delivers, nor does it know if the data was incorrectly transmitted.
This new system proved itself more flexible in dealing with the quickly growing ARPANET. By 1973, international connections had been established with NORSAR in Norway and with the University College in London. By 1977, the ARPANET had connected with other networks, such as the Ethernet. The resultant global network became known as the "Internet". When the ARPANET was decommissioned in 1989, the Internet was a vastly different place than it had been twenty years earlier.
At this point, the Internet was still an esoteric, although tremendously useful, agora of science and engineering. Several steps helped to transform it into a populist, free-for-all of ideas, goods, media and information.
In 1984, the DNS (Domain Name System) was introduced. The DNS is essentially a large lookup table that correlates an English name with a machine name. For example, if a user wants to talk with eesn9.ews.uiuc.edu, the lookup table will know that he really wants to talk to the computer with the address of 130.126.161.193. The DNS system connecting to other machines (especially those you hadn't used before) far easier, in the five years after 1984, the number of hosts increased a hundred-fold, to 100,000. By 1989, the Internet was a truly international network, with nodes in Australia (au), Germany (de), Israel (il), Italy (it), Japan (jp), Mexico (mx), the Netherlands (nl), New Zealand (nz), and the United Kingdom (uk).
The invention which really led to the public explosion of the Internet came in 1991, when Tim Berners-Lee, working at CERN, released the code for the World Wide Web (WWW). The web offered a substantially better method of sharing information and documentation than did the older ftp and telnet protocols. In 1992, the number of hosts reached 1,000,000, and Jean Armour Polly coined the phrase "surfing the Internet".
But the web was still a text-based phenomenon. In 1993, however, Mosaic, developed here at the UI, was released, bringing a graphical medium to the web. Mosaic was an instant hit, and the resultant firestorm of interest caused the web to grow by 341,634% that year. By 1994, you could order from Pizza Hut over the Internet. Other commercial entities included banks, shopping malls, radio stations and unfortunatley, lawyers. The web overtook telnet, and was second only to ftp in terms of bandwidth usage on the Internet.
As popular usage of the Internet surged, commercial interests quickly moved in. Entreprenuers were soon hawking goods and services, coexisting quite well with more traditional scientific pursuits. There are already two obvious trends in the development of the Internet. These are the rise of e-commerce and information deflation.
The rise of e-commerce has already transformed society. This much is obvious. Anyone can buy electronics, books, or anything else, from just about anywhere, tax-free. This has opened up competition, and will drive many older businesses under. Even the august Encyclopedia Brittanica is no longer the pricy, leather book set. All its information is freely available online.
But the rise of the e-commerce is going to change the face of society in ways that are not foreseeable right now. Our politicians like to prattle on about taxing the Internet. But many of them are just now realizing how little control they can actually exert. With the advent of unbreakable encryption routines, there is nothing to prevent the emergence, for example, of an online currency, that could be used to buy goods and services across the net. Such a system, outside the bounds of the traditional banking system, and not subject to governmental regulation, could cause significant loses of government revenue.
In any event, the e-commerce boom will increase the scale on which businesses, even small ones, operate. The small businessman has far more opportunity in cyberspace than he did in the store on the corner. The only real stumbling block is name recognization. And on the web, name recognition is quickly gained, and even more quickly lost. There are lots of peopll out there willing to tell everyone who will listen exactly how good or how bad the service was. Taken collectively, these voices have a lot of weight with shoppers.
I imagine that e-commerce will increase both the availability and the quality of goods and services throughout the US, as well as the world.
The other marvel of the information age is the remarkable deflation in the cost of information. The price of information is dropping at an unbelievable rate. The computer network that allowed researchers and scientists to share their data and information has also made that information available to everyone.
No one would have believed you, ten years ago, if you told him that he could red for free every major US and British newspaper, that he could read for free the entire Oxford English Dictionary and the Encyclopedia Britannica, that he could download for free all the raw data from the Hubble Space Telescope, or, that, in 5 minutes of "surfing the web", find good information on just about any topic he could possibly imagine, that he could listen to thousands of radio stations from all across the world, that he could talk to anyone, anywhere for the price of a local phone call, that he could buy just about anything he wanted, cheaper and quicker, and tax-free.
If I were he, I would have laughed, and called my future self an irrepressible optimist, hopelessly out of touch with the "laws of economics". But I would have been wrong. And, thanks to the success of the open source movement, even the software and technology behind this awesome advance are free. Economists always claimed that the "invisible hand of the market" would force profit to zero, but that never turned out to be the case. Now, in the high-tech world of computers and information, this may eventually come to pass. If knowledge liberates, if the truth sets one free, then one can argue that we have our freedom. Now we must keep it.